PhD Forum: Enabling Autonomic IoT for Smart Urban Services

12/24/2019 ∙ by Muhammad Junaid Farooq, et al. ∙ 0

The development of autonomous cyber-physical systems (CPS) and advances towards the fifth generation (5G) of wireless technology is promising to revolutionize many industry verticals such as Healthcare, Transportation, Energy, Retail Services, Building Automation, Education, etc., leading to the realization of the smart city paradigm. The Internet of Things (IoT), enables powerful and unprecedented capabilities for intelligent and autonomous operation. We leverage ideas from Network Science, Optimization Decision Theory, Incentive Mechanism Design, and Data Science/Machine Learning to achieve key design goals, in IoT-enabled urban systems, such as efficiency, security resilience, and economics.

READ FULL TEXT VIEW PDF
POST COMMENT

Comments

There are no comments yet.

Authors

This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.

I Introduction

The cyber-physical integration of devices in the IoT is enabling the development of a myriad of applications and services. At each layer of interaction between the systems, there are decision problems that arise for achieving various different objectives. Effective decision-making at different fronts is essential to enhance the efficiency, economics, security, and resilience of these systems. This research focuses on developing cyber-physical decision mechanisms. For instance, at the device level, it is important to ensure wireless connectivity and communication. If the communication infrastructure is in place, there is a need for spectrum allocation and reservation. However, in case of challenged infrastructure, the connectivity might be achieved using an overlay network of aerial base stations [1] [2]. Decisions have to be made on allocation of spectrum to users and configuration/placement of aerial base stations to provide adequate coverage and connectivity. Once the connectivity is achieved, the networks are used for information dissemination [3]. Furthermore, there is a need to design security mechanisms against stealthy adversarial threats that may be using the same communication networks to infiltrate and sabotage network operation. The next frontier is the use of cloud computing resources by smart devices. It is important to efficiently allocate and price the computational resources in order to provide a high quality of experience to the users and to generate high revenues for the cloud/fog service provider. Finally, at the application front, there are scenarios where resource provisioning decisions need to be made for service requests that appear randomly in space and time as in instance in the case of smart urban environments.

Fig. 1: Cross-layer decision-making in the IoT ecosystem.

Ii Challenges

Dynamic mechanism design traditionally has been focused on principal-agent type of models with a mechanism designer and participants. In IoT systems, these mechanisms may not always be available. The IoT provides new avenues for designing policies and mechanisms in a dynamic setting at multiple different levels. For instance, at the physical layer, there is a need for connectivity mechanisms while at the higher layers, policy decisions such as allocation and pricing are required. Cyber-Physical coupling needs to be explored in the realm of the IoT which is non-existent in traditional communication and computing systems. Moreover, the IoT ecosystem is inherently a large-scale, complex, and dynamic. Therefore, centralized mechanisms are difficult to implement and are in-feasible. Therefore, more distributed approaches are required.

Iii Contributions

Autonomous operation of CPS/IoT systems requires an interdisciplinary and cross-layer approach due to the coupling between cyber and physical components. The state and communication at the cyber layer influences the dynamics and control at the physical layer and vice versa. An high level illustration of the system is provided in Fig. 1. The following subsections provide details on some of the key thrusts of this research.

Iii-a Autonomic Networked CPS: From Military to Civilian Applications

Connected CPS networks are used for dissemination of data and information for enhanced situational awareness and decision-making. Some example applications are illustrated in Fig. 2. Of particular interest are the military and tactical networks with stringent requirements such as support for extreme heterogeneity, rapid re-configurability, and mosaic warfare needs. In this respect, one of the key contributions of our research is the development of a secure and re-configurable network design framework suitable for adversarial environments such as the Internet of Battlefield Things (IoBT) [4][5]. Networks with enhanced data dissemination capabilities also open doors for malicious activity and malware. Furthermore, malicious entities such as compromised supply chain actors may exploit backdoor channels for stealthy takeover and cause large scale coordinated attacks. To tackle this security risk posed by wireless connectivity proliferation in IoT, we developed a mechanism for inspecting devices based on their wireless neighbourhood that prevents a large scale coordinated attack while causing minimum operational interruption [6].

Fig. 2: Autonomic CPS/IoT Systems.

Iii-B Strategic Resource Provisioning for Mission-Critical IoT Services

At the communication layer of the IoT, the spectrum resources needs to be effectively provisioned to applications according to their performance requirements and power limitations [7]. Similarly, at the cloud layer, computing nodes and data processing resources need to be allocated and priced strategically to ensure maximum revenue for the cloud service provider. In this regard, we have developed a real-time resource allocation and pricing framework for cloud-enabled IoT systems, where the computational complexity of arriving tasks is evaluated and is accordingly assigned to the available computing bundles [8]. Similarly, for low latency applications, a framework is developed to appropriately select the resources available at one of the fog/edge computing nodes for real-time tasks [9]. We also extend the allocation and pricing framework to the case where there is uncertainty in the spatial domain in addition to the temporal domain [10].

Iii-C Data-Driven Decisions for Urban Service Provisioning

IoT-driven urban services are rapidly emerging in almost all industry verticals. Companies like Uber, lyft, Via, etc. have come up with a range of on-demand urban mobility solutions and are moving towards autonomous microtransit solutions. Similarly, there is a strong interest in autonomous vehicle and drone based delivery services. These applications require intelligent decision-making by autonomous agents. Emergency response and first responder tactical systems are key to the safety and resilience of smart cities. The use of sensing capabilities, historical data, and distributed data processing can not only assist early detection and rapid response, but can also help in emergency preparedness. For instance, wildfires are natural disasters that pose a significant threat to the metropolis in terms of life, health, safety, etc. The use of data from sensors reporting temperatures along with weather data can help in determining the timing and location of such incidents. Similarly, timely emergency response is critical for urban safety and quality of life. Some example urban services are depicted in Fig. 3.

Fig. 3: Examples of urban on-demand services.

Iv Conclusion

This research is an attempt to lay the theoretical foundations of decision science in IoT network design and operation. It leverages tools and theories from various different systems sciences such as mathematical epidemiology, spatial point processes, stochastic processes, optimal control theory, and optimization to tackle these challenging problems. It addresses the challenges and problems at various different levels across the IoT stack. We hope that this work will form the basis for the development of a comprehensive science for decision mechanisms in the IoT networks domain.

References

  • [1] M. J. Farooq and Q. Zhu, “A multi-layer feedback system approach to resilient connectivity of remotely deployed mobile internet of things,” IEEE Transactions on Cognitive Communications and Networking, vol. 4, no. 2, pp. 422–432, June 2018.
  • [2] ——, “Cognitive connectivity resilience in multi-layer remotely deployed mobile internet of things,” in GLOBECOM 2017 - 2017 IEEE Global Communications Conference, Dec 2017, pp. 1–6.
  • [3] M. J. Farooq, H. ElSawy, Q. Zhu, and M. Alouini, “Optimizing mission critical data dissemination in massive iot networks,” in 2017 15th International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks (WiOpt), May 2017, pp. 1–6.
  • [4] M. J. Farooq and Q. Zhu, “On the secure and reconfigurable multi-layer network design for critical information dissemination in the internet of battlefield things (iobt),” IEEE Transactions on Wireless Communications, vol. 17, no. 4, pp. 2618–2632, April 2018.
  • [5] ——, “Secure and reconfigurable network design for critical information dissemination in the internet of battlefield things (iobt),” in 2017 15th International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks (WiOpt), May 2017, pp. 1–8.
  • [6] ——, “Modeling, analysis, and mitigation of dynamic botnet formation in wireless iot networks,” IEEE Transactions on Information Forensics and Security, vol. 14, no. 9, pp. 2412–2426, Sep. 2019.
  • [7] ——, “Optimal dynamic contract for spectrum reservation in mission-critical unb-iot systems,” in 2018 16th International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks (WiOpt), May 2018, pp. 1–6.
  • [8] ——, “Adaptive and resilient revenue maximizing dynamic resource allocation and pricing for cloud-enabled iot systems,” in 2018 Annual American Control Conference (ACC), June 2018, pp. 5292–5297.
  • [9] ——, “QoE based revenue maximizing dynamic resource allocation and pricing for fog-enabled mission-critical iot applications,” under revision in IEEE Transactions on Mobile Computing, 2019.
  • [10] ——, “Dynamic spatio-temporal resource provisioning for on-demand urban services in smart cities,” under revision in IEEE Transactions on Cognitive Communications and Networking, 2019.